You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi,
I'm wondering the purpose of using two same negative prompt embedding. prompt_embeds = torch.cat([prompt_embeds, negative_prompt_embeds, negative_prompt_embeds])
And you set noise_pred = ( noise_pred_uncond + self.guidance_scale * (noise_pred_text - noise_pred_image) + self.image_guidance_scale * (noise_pred_image - noise_pred_uncond) )
In this case why not merge the scale factor?
The text was updated successfully, but these errors were encountered:
Hi,
I'm wondering the purpose of using two same negative prompt embedding.
prompt_embeds = torch.cat([prompt_embeds, negative_prompt_embeds, negative_prompt_embeds])
And you set
noise_pred = ( noise_pred_uncond + self.guidance_scale * (noise_pred_text - noise_pred_image) + self.image_guidance_scale * (noise_pred_image - noise_pred_uncond) )
In this case why not merge the scale factor?
The text was updated successfully, but these errors were encountered: